Systems Biology Approach Predicts Antibody Signature Associated

ARTICLE pubs.acs.org/jpr

Systems Biology Approach Predicts Antibody Signature Associated with Brucella melitensis Infection in Humans Li Liang,*,† Xiaolin Tan,† Silvia Juarez,† Homarh Villaverde,‡ Jozelyn Pablo,† Rie Nakajima-Sasaki,† Eduardo Gotuzzo,‡ Mayuko Saito,#,r Gary Hermanson,§ Douglas Molina,§ Scott Felgner,§ W. John W. Morrow,§ Xiaowu Liang,§ Robert H. Gilman,||,z,r D. Huw Davies,† Ren
ee M. Tsolis,^ Joseph M. Vinetz,#,z and Philip L. Felgner†,* †

Department of Medicine, Division of Infectious Diseases, University of California, Irvine, California 92697, United States Alexander von Humboldt Institute of Tropical Medicine, Universidad Peruana Cayetano Heredia, Lima, Peru § Antigen Discovery, Inc., Irvine, California 92618, United States Department of International Health, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland 21205, United States ^ Department of Medical Microbiology and Immunology, University of California, Davis, Davis, California 95616, United States # Division of Infectious Diseases, Department of Medicine, University of California San Diego School of Medicine, La Jolla, California 92093-0741, United States r Asociacion Benefica PRISMA, Lima, Peru z Department of Microbiology, Universidad Peruana Cayetano Heredia, Lima, Peru

)

‡

bS Supporting Information ABSTRACT: A complete understanding of the factors that determine selection of antigens recognized by the humoral immune response following infectious agent challenge is lacking. Here we illustrate a systems biology approach to identify the antibody signature associated with Brucella melitensis (Bm) infection in humans and predict proteomic features of serodiagnostic antigens. By taking advantage of a full proteome microarray expressing previously cloned 1406 and newly cloned 1640 Bm genes, we were able to identify 122 immunodominant antigens and 33 serodiagnostic antigens. The reactive antigens were then classified according to annotated functional features (COGs), computationally predicted features (e.g., subcellular localization, physical properties), and protein expression estimated by mass spectrometry (MS). Enrichment analyses indicated that membrane association and secretion were significant enriching features of the reactive antigens, as were proteins predicted to have a signal peptide, a single transmembrane domain, and outer membrane or periplasmic location. These features accounted for 67% of the serodiagnostic antigens. An overlay of the seroreactive antigen set with proteomic data sets generated by MS identified an additional 24%, suggesting that protein expression in bacteria is an additional determinant in the induction of Brucella-specific antibodies. This analysis indicates that one-third of the proteome contains enriching features that account for 91% of the antigens recognized, and after B. melitensis infection the immune system develops significant antibody titers against 10% of the proteins with these enriching features. This systems biology approach provides an empirical basis for understanding the breadth and specificity of the immune response to B. melitensis and a new framework for comparing the humoral responses against other microorganisms. KEYWORDS: Brucella melitensis, enriching features, immune response, protein microarray

’ INTRODUCTION A major component of the adaptive immune response to infection is the generation of protective and long-lasting humoral immunity, but factors governing selection of the particular antigens recognized are unclear.1,2 It is not uncommon for viruses encoding a small number of proteins to generate antibodies against each encoded protein. But for infectious agents containing hundreds or thousands of proteins only a subset of the proteome is recognized and little is known about the extent or the r 2011 American Chemical Society

characterisitics of this subset of antigens. Methods for making a complete empirical accounting of the immunoproteome have limitations, particularly when the genome of the organism is large. Here we describe a B. melitensis proteome microarray that enables this problem to be directly addressed by applying an unbiased systems biology approach to identify immunodominant Received: June 30, 2011 Published: August 24, 2011 4813

dx.doi.org/10.1021/pr200619r | J. Proteome Res. 2011, 10, 4813–4824

Journal of Proteome Research and serodiagnostic antigens and to classify the reactive antigens based on functional and physical properties. Protein microarrays can be used with relative ease to probe the entire proteome of different infectious microorganisms including bacteria, viruses and parasites.313 This approach permits assessing the repertoire of antibodies produced in response to infections or vaccinations from large collections of individual patient sera, and can be used to perform large-scale sero-epidemiological and serosurveillance analyses not possible with other technologies, while consuming small quantities ( 0.05) antigens to the right. Shown are the 33 serodiagnostic protein antigens, Bm LPS, and a subset of cross-reactive antigens.

were sequenced and >99% of sequenced clones had the correct sequence. All 3046 Bm ORFs cloned in pXT7 vector (>95% of proteome) were expressed under T7 promoter in the E. coli in vitro transcription/translation system, and printed on microarrays. Over 98% of protein spots were confirmed positive for expression (Figure S1A, Supporting Information). Human Antibody Profile

Bm protein arrays were probed with sera from acute brucellosis patients in Lima, Peru obtained within 13 weeks of the onset of symptoms, and sera from Bm culture-positive humans (Figure S1B, Supporting Information) showed robust reactivity against a collection of antigens compared to unexposed individuals. Among the 3046 antigens tested, 1464 antigens reacted with at least one culture positive individual, accounting for 48% of the proteome (Figure 1). Within this immunoproteome, 122

protein antigens were defined “serodominant”, with mean reactivity greater than the mean of the no DNA controls plus 2.5 standard deviations among culture positive individuals (Figure 2, Figure S2 and Table S1, Supporting Information). Of these, 33 protein antigens were serodiagnostic, and were significantly differentially reactive between na€ive and culture positive patients from Peru (Benjamini and Hochberg adjusted Cyber-T p-value 0.7, pSort Outermembrane, pSort Periplasmic, or pI < 5; and (3) Protein expression as detected by mass spectrometry. Numbers of overlapping antigens for two or three categories are shown in Figure 6.

were printed on our proteome chip. Not surprisingly, expression was a significant enriching feature with 5.2-fold enrichment among serodiagnostic antigens (p-value 1.87  1011), and 2.9fold in serodominant antigens (p-value 5.28  1011) (Table 1). Interestingly, the fold enrichment was directly proportional to the number of peptides detected (Figure 5C). There was 8.4- fold enrichment of serodiagnostic antigens in proteins detected with more than 1 peptide (p-value 3.56  1012), 16.2- fold in proteins detected with 5 or more peptides (p-value 1.30  107), 23.1- fold enrichment in proteins detected with 7 or more peptides (p-value 1.90  105), and 20.5-fold enrichment in proteins detected with 9 or more peptides (p-value 3.91  103). Conversely, proteins that were not identified by MS were significantly underrepresented at 0.4-fold among the serodiagnostic and 0.8-fold among serodominant antigens. The data suggests that protein expression level is an important factor contributing to antigenicity of protein antigens. This was also supported by data in other experimental systems from different research groups.55 There are a total of 10 enriching features that fall into 3 categories: (i) functionally annotated COGs U, M, N and O, (ii) computationally predicted features (TMHMM = 1, SignalP > 0.7,

pSort Outermembrane, pSort Periplasmic, and pI < 5), and (iii) MS evidence of expression. The Venn diagrams summarize the number of enriched proteins found in each of these 3 categories, showing overlap among the proteins found in each category (Figure 6). There are 338 proteins in the enriched COG categories, 696 computationally predicted proteins, and 356 proteins that were positive by MS (Table 2). Accounting for the overlap between categories, 1128 proteins in all enriched categories represented 37% of the proteome. The microarray results empirically identified 33 serodiagnostic antigens and 89 cross-reactive antigens. COGs accounted for 30% of the serodiagnostic hits, the computationally predicted features accounted for 61% of the hits, and MS positive proteins contained 61% of the hits. All together the three categories account for 37% of the proteome and include 91% of the serodiagnostic antigens. But there is only 1 unique serodiagnostic COG antigen that is not represented in either the computationally predicted or MS+ categories (BMEI1060, Table S1, Supporting Information). So by combining the pool of computationally predicted and MS+ proteins representing 30% of the proteome, 88% of the serodiagnostic antigens can be predicted. 4820

dx.doi.org/10.1021/pr200619r |J. Proteome Res. 2011, 10, 4813–4824

Journal of Proteome Research

’ DISCUSSION Today with increasing efficiency, accuracy and speed we access completed genome sequences from thousands of infectious microorganisms. However, systems- level studies to understand the complete network of antibody responses have not been widely performed. This work was motivated by the disturbing lack of a detailed scientific understanding of the antibody responses to infectious diseases. We have been taking a systems biology approach to account for all of the antibodies that develop after exposure to infectious microorganisms and to identify specific antibody signatures associated with each disease, in order to understand the molecular basis for antigen selection by the immune system and to predict serodiagnostic and vaccine antigens targets. This study is the first full proteome-wide serological analysis of B. melitensis. Our previous study identified a set of serodiagnostic antigens from a randomly selected half of the B. melitensis proteome. However, an important difference between the present study and this previous report is that analysis of the complete proteome not only has allowed us to delineate additional serodiagnostic antigens but, more fundamentally, also provided the basis for a rigorous, comprehensive and quantitative determination of basic biological characteristics of the entire set of the serodominant antigens on a genomic level. The current standard serological screening assay, an agglutination test that uses tinted, killed bacteria as antigen (Rose Bengal) is based primarily on identification of antibodies to LPS in patient serum. In the present microarray study, we confirmed LPS reactivity (with purified LPS spotted onto the microarray), and identified novel protein antigens. The signature antigens were validated using traditional Western blots. Although antibody responses against all 33 protein antigens used together can predict disease with 92% accuracy, using the serological responses against the top 3 individual antigens together improves accuracy to >98% for diagnosis of acute human brucellosis. One limitation is that the humans studied here only had acute or subacute brucellosis-fever but not focal or chronic disease. Future studies are needed to assess suspected cases that are agglutination test negative, for example chronic neurobrucellosis or focal disease such as orchitis or vertebral osteomyelitis. Eleven of the 33 serodiagnostic antigens identified in this study were also identified on the pilot array5 (Table S1, Supporting Information). Our findings are in good agreement with published studies that identified well characterized antigens from other Brucella spp, including Bp26 (BMEI0536),20,25,31,56 HtrA/ DegP (BMEI1330),28 Omp16 (BMEI0340),29 the chaperonin GroEL protein (BMEII1048) and Omp 10 (BMEII0017).21,57 In addition to the well characterized antigens, we also identified 21 novel serodiagnostic antigens on the current array, including top antigen hypothetical protein BMEI0805. Two pyruvate dehydrogenase complex molecules and the associated acetyl CoA hydrolase, which together form a large multimeric structure consisting of 60 subunits, are also differentially recognized; this large enzyme complex is found in most bacteria and is frequently a target of immune recognition for other infections.6,58 Protein microarrays enable enrichment analyses to identify proteomic features that are enriched in the immunodominant antigen set, and development of protein antigenicity prediction tools based on enriched features. The prediction tools then can be applied on a high-throughput scale to existing or new proteomes to identify key antigenic proteins that may have serodiagnostic or protective characteristics. Enrichment analysis identified 10 proteomic

ARTICLE

features that are enriched in the serodiagnostic signature antigens. The proteomic features fall into 3 categories: (i) COG annotations, (ii) computationally predicted proteomic features, and (iii) proteins with MS evidence of expression in viable organisms. No single proteomic feature or category of features is sufficient to identify all these signature antigens. Our data suggests that cloning of 37% of genome with these enriching features would reveal more than 90% of serodiagnostic antigens. We have classified the reactive immunodominant antigens for numerous agents and have consistently found these features predict antigenicity.6,7,10,59 These results describe the relationship between antigenicity and in vivo expression of individual proteins. In our study, we utilized the expression of proteins quantified by MS of a closely related strain, B. abortus during in vitro growth. We found that mass spec positive antigens were significantly enriched among serodiagnostic antigens but not among the cross- reactive antigens. This validates the classification of the serodiagnostic antigens being derived from actively replicating organisms, and validates the classification of cross-reactive antigens being derived from previous exposure to nonbrucellosis infections. Although the number of peptides observed per protein has been applied to estimating protein abundance,6062 we are aware that mass spectra of cultured organisms may not reflect the expression during infection in vivo. First, nutrients are not limiting in log phase growth in vitro, whereas in vivo, Brucella is likely to be within a nutrient-limited intracellular environment.63 Second, the MS method may not be particularly quantitative, especially for membrane proteins. Our protein array technology is also able to provide strong evidence of the comprehensive set of proteins expressed in vivo within a mammalian host by B. melitensis, by virtue of their exposure to the host immune system. The expression and abundance of proteins during in vivo growth merits further examination by other novel technologies, such as measurement of transcript abundance from RNA-seq (“deep sequencing”) technique.64 The B. melitensis immunoproteome comprises 1464 antigens that are significantly reactive in at least one of the individuals in this study. Only 122 serodominant antigens are significantly recognized by most of the individuals. Answers to questions of why and how the immune system focuses on 4% of the potential target antigens are not yet apparent from this work. One might expect that an immune response against a larger collection of antigens could result in a more effective immune response attack against the infectious agent. But antibody responses against thousands of antigens from hundreds of clinical infections could accumulate during a lifetime, leading to cross reactivity against autologous antigens and autoimmune chaos. This could provide evolutionary selection pressure favoring a more focused response to infection. The observation that dozens of organism- specific antibody responses develop after B. melitensis infection is consistent with similar observations from other infectious agents, and has implications for subunit vaccine discovery and development. Mimicking the natural response to infection could be considered a viable strategy for vaccine development but most subunit vaccines aim to derive protection from immunization with only a single antigen. Attenuated or killed whole organism vaccines produce an antibody reactivity profile against dozens of antigens more similar to natural infection.6 This systematic genome scale analysis of human antibody responses against B. melitensis proteins provides a top hit list of antigens worthy of assessing for improved diagnostics, and 4821

dx.doi.org/10.1021/pr200619r |J. Proteome Res. 2011, 10, 4813–4824

Journal of Proteome Research furthermore, enables development of a predictive model of proteomic features that determine whether a protein is antigenic and produces antibodies that confer protection. This systems biology approach provides an empirical basis for understanding the breadth and specificity of the immune response to B. melitensis and a new framework for comparing the humoral responses against other organisms.

’ ASSOCIATED CONTENT

bS

Supporting Information Supplementary tables and figures. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Li Liang ([email protected]) or Philip L. Felgner (pfelgner@ uci.edu). Phone: 949-824-1407. Fax: 949-824- 0481. Notes

Conflict of InterestP.L.F. and D.H.D. have patent applications pertaining to this work and own stock in Anitigen Discovery Inc. that has licensed the technology. X.L. is an inventor on these patent applications and is employed at Anitigen Discovery Inc. W.J.W.M. is employed at Anitigen Discovery Inc.

’ ACKNOWLEDGMENT The Brucella melitensis lipopolysaccharide was kindly provided by Kailash Patra in Joseph Vinetz lab. This work was partially supported by grants from the U.S. Public Health Service, National Institutes of Health 1U01AI075420, 1K24AI068903, and D43TW007120 (J.M.V.), U01AI078213 and U54AI065359 (P.L.F.), and 5R44AI058365-05 (D.H.D.). ’ REFERENCES (1) Mayers, C.; Duffield, M.; Rowe, S.; Miller, J.; Lingard, B.; Hayward, S.; Titball, R. W. Analysis of known bacterial protein vaccine antigens reveals biased physical properties and amino acid composition. Comp. Funct. Genomics 2003, 4 (5), 468–78. (2) Rappuoli, R. Reverse vaccinology, a genome-based approach to vaccine development. Vaccine 2001, 19 (1719), 2688–91. (3) Kunnath-Velayudhan, S.; Salamon, H.; Wang, H. Y.; Davidow, A. L.; Molina, D. M.; Huynh, V. T.; Cirillo, D. M.; Michel, G.; Talbot, E. A.; Perkins, M. D.; Felgner, P. L.; Liang, X.; Gennaro, M. L. Dynamic antibody responses to the Mycobacterium tuberculosis proteome. Proc. Natl. Acad. Sci. U.S.A. 2010, 107 (33), 14703–8. (4) Crompton, P. D.; Kayala, M. A.; Traore, B.; Kayentao, K.; Ongoiba, A.; Weiss, G. E.; Molina, D. M.; Burk, C. R.; Waisberg, M.; Jasinskas, A.; Tan, X.; Doumbo, S.; Doumtabe, D.; Kone, Y.; Narum, D. L.; Liang, X.; Doumbo, O. K.; Miller, L. H.; Doolan, D. L.; Baldi, P.; Felgner, P. L.; Pierce, S. K. A prospective analysis of the Ab response to Plasmodium falciparum before and after a malaria season by protein microarray. Proc. Natl. Acad. Sci. U.S.A. 2010, 107 (15), 6958–63. (5) Liang, L.; Leng, D.; Burk, C.; Nakajima-Sasaki, R.; Kayala, M. A.; Atluri, V. L.; Pablo, J.; Unal, B.; Ficht, T. A.; Gotuzzo, E.; Saito, M.; Morrow, W. J.; Liang, X.; Baldi, P.; Gilman, R. H.; Vinetz, J. M.; Tsolis, R. M.; Felgner, P. L. Large scale immune profiling of infected humans and goats reveals differential recognition of Brucella melitensis antigens. PLoS Negl. Trop. Dis. 2010, 4 (5), e673. (6) Eyles, J. E.; Unal, B.; Hartley, M. G.; Newstead, S. L.; FlickSmith, H.; Prior, J. L.; Oyston, P. C.; Randall, A.; Mu, Y.; Hirst, S.; Molina, D. M.; Davies, D. H.; Milne, T.; Griffin, K. F.; Baldi, P.; Titball,

ARTICLE

R. W.; Felgner, P. L. Immunodominant Francisella tularensis antigens identified using proteome microarray. Proteomics 2007, 7 (13), 2172–83. (7) Felgner, P. L.; Kayala, M. A.; Vigil, A.; Burk, C.; Nakajima-Sasaki, R.; Pablo, J.; Molina, D. M.; Hirst, S.; Chew, J. S.; Wang, D.; Tan, G.; Duffield, M.; Yang, R.; Neel, J.; Chantratita, N.; Bancroft, G.; Lertmemongkolchai, G.; Davies, D. H.; Baldi, P.; Peacock, S.; Titball, R. W. A Burkholderia pseudomallei protein microarray reveals serodiagnostic and cross-reactive antigens. Proc. Natl. Acad. Sci. U.S.A. 2009, 106 (32), 13499–504. (8) Beare, P. A.; Chen, C.; Bouman, T.; Pablo, J.; Unal, B.; Cockrell, D. C.; Brown, W. C.; Barbian, K. D.; Porcella, S. F.; Samuel, J. E.; Felgner, P. L.; Heinzen, R. A. Candidate antigens for Q fever serodiagnosis revealed by immunoscreening of a Coxiella burnetii protein microarray. Clin. Vaccine Immunol. 2008, 15 (12), 1771–9. (9) Barbour, A. G.; Jasinskas, A.; Kayala, M. A.; Davies, D. H.; Steere, A. C.; Baldi, P.; Felgner, P. L. A genome-wide proteome array reveals a limited set of immunogens in natural infections of humans and whitefooted mice with Borrelia burgdorferi. Infect. Immun. 2008, 76 (8), 3374–89. (10) Vigil, A.; Ortega, R.; Jain, A.; Nakajima-Sasaki, R.; Tan, X.; Chomel, B. B.; Kasten, R. W.; Koehler, J. E.; Felgner, P. L. Identification of the feline humoral immune response to Bartonella henselae infection by protein microarray. PLoS One 2010, 5 (7), e11447. (11) Driguez, P.; Doolan, D. L.; Loukas, A.; Felgner, P. L.; McManus, D. P. Schistosomiasis vaccine discovery using immunomics. Parasit Vectors 2010, 3, 4. (12) Davies, D. H.; Molina, D. M.; Wrammert, J.; Miller, J.; Hirst, S.; Mu, Y.; Pablo, J.; Unal, B.; Nakajima-Sasaki, R.; Liang, X.; Crotty, S.; Karem, K. L.; Damon, I. K.; Ahmed, R.; Villarreal, L.; Felgner, P. L. Proteome-wide analysis of the serological response to vaccinia and smallpox. Proteomics 2007, 7 (10), 1678–86. (13) Luevano, M.; Bernard, H. U.; Barrera-Saldana, H. A.; Trevino, V.; Garcia-Carranca, A.; Villa, L. L.; Monk, B. J.; Tan, X.; Davies, D. H.; Felgner, P. L.; Kalantari, M. High-throughput profiling of the humoral immune responses against thirteen human papillomavirus types by proteome microarrays. Virology 2010, 405 (1), 31–40. (14) Pappas, G.; Akritidis, N.; Bosilkovski, M.; Tsianos, E. Brucellosis. N. Engl. J. Med. 2005, 352 (22), 2325–36. (15) Young, E. J. An overview of human brucellosis. Clin. Infect. Dis 1995, 21 (2), 283–9; quiz 290. (16) Franco, M. P.; Mulder, M.; Gilman, R. H.; Smits, H. L. Human brucellosis. Lancet Infect. Dis. 2007, 7 (12), 775–86. (17) Mantur, B. G.; Amarnath, S. K.; Shinde, R. S. Review of clinical and laboratory features of human brucellosis. Indian J. Med. Microbiol. 2007, 25 (3), 188–202. (18) Pappas, G.; Papadimitriou, P.; Akritidis, N.; Christou, L.; Tsianos, E. V. The new global map of human brucellosis. Lancet Infect. Dis. 2006, 6 (2), 91–9. (19) Al Dahouk, S.; Jubier-Maurin, V.; Scholz, H. C.; Tomaso, H.; Karges, W.; Neubauer, H.; Kohler, S. Quantitative analysis of the intramacrophagic Brucella suis proteome reveals metabolic adaptation to late stage of cellular infection. Proteomics 2008, 8 (18), 3862–70. (20) Cloeckaert, A.; Debbarh, H. S.; Vizcaino, N.; Saman, E.; Dubray, G.; Zygmunt, M. S. Cloning, nucleotide sequence, and expression of the Brucella melitensis bp26 gene coding for a protein immunogenic in infected sheep. FEMS Microbiol. Lett. 1996, 140 (23), 139–44. (21) Connolly, J. P.; Comerci, D.; Alefantis, T. G.; Walz, A.; Quan, M.; Chafin, R.; Grewal, P.; Mujer, C. V.; Ugalde, R. A.; DelVecchio, V. G. Proteomic analysis of Brucella abortus cell envelope and identification of immunogenic candidate proteins for vaccine development. Proteomics 2006, 6 (13), 3767–80. (22) Gor, D.; Mayfield, J. E. Cloning and nucleotide sequence of the Brucella abortus groE operon. Biochim. Biophys. Acta 1992, 1130 (1), 120–2. (23) Lamontagne, J.; Forest, A.; Marazzo, E.; Denis, F.; Butler, H.; Michaud, J. F.; Boucher, L.; Pedro, I.; Villeneuve, A.; Sitnikov, D.; Trudel, K.; Nassif, N.; Boudjelti, D.; Tomaki, F.; Chaves-Olarte, E.; Guzman-Verri, C.; Brunet, S.; Cote-Martin, A.; Hunter, J.; Moreno, E.; 4822

dx.doi.org/10.1021/pr200619r |J. Proteome Res. 2011, 10, 4813–4824

Journal of Proteome Research Paramithiotis, E. Intracellular adaptation of Brucella abortus. J. Proteome Res. 2009, 8 (3), 1594–609. (24) Letesson, J. J.; Tibor, A.; van Eynde, G.; Wansard, V.; Weynants, V.; Denoel, P.; Saman, E. Humoral immune responses of Brucellainfected cattle, sheep, and goats to eight purified recombinant Brucella proteins in an indirect enzyme-linked immunosorbent assay. Clin. Diagn. Lab. Immunol. 1997, 4 (5), 556–64. (25) Lindler, L. E.; Hadfield, T. L.; Tall, B. D.; Snellings, N. J.; Rubin, F. A.; Van De Verg, L. L.; Hoover, D.; Warren, R. L. Cloning of a Brucella melitensis group 3 antigen gene encoding Omp28, a protein recognized by the humoral immune response during human brucellosis. Infect. Immun. 1996, 64 (7), 2490–9. (26) Mayfield, J. E.; Bricker, B. J.; Godfrey, H.; Crosby, R. M.; Knight, D. J.; Halling, S. M.; Balinsky, D.; Tabatabai, L. B. The cloning, expression, and nucleotide sequence of a gene coding for an immunogenic Brucella abortus protein. Gene 1988, 63 (1), 1–9. (27) Rolan, H. G.; den Hartigh, A. B.; Kahl-McDonagh, M.; Ficht, T.; Adams, L. G.; Tsolis, R. M. VirB12 is a serological marker of Brucella infection in experimental and natural hosts. Clin. Vaccine Immunol. 2008, 15 (2), 208–14. (28) Roop, R. M., 2nd; Fletcher, T. W.; Sriranganathan, N. M.; Boyle, S. M.; Schurig, G. G. Identification of an immunoreactive Brucella abortus HtrA stress response protein homolog. Infect. Immun. 1994, 62 (3), 1000–7. (29) Tibor, A.; Weynants, V.; Denoel, P.; Lichtfouse, B.; De Bolle, X.; Saman, E.; Limet, J. N.; Letesson, J. J. Molecular cloning, nucleotide sequence, and occurrence of a 16.5-kilodalton outer membrane protein of Brucella abortus with similarity to pal lipoproteins. Infect. Immun. 1994, 62 (9), 3633–9. (30) Yang, X.; Hudson, M.; Walters, N.; Bargatze, R. F.; Pascual, D. W. Selection of protective epitopes for Brucella melitensis by DNA vaccination. Infect. Immun. 2005, 73 (11), 7297–303. (31) Rossetti, O. L.; Arese, A. I.; Boschiroli, M. L.; Cravero, S. L. Cloning of Brucella abortus gene and characterization of expressed 26kilodalton periplasmic protein: potential use for diagnosis. J. Clin. Microbiol. 1996, 34 (1), 165–9. (32) Baker, P. J.; Wilson, J. B. Hypoferremia in mice and its application to the bioassay of endotoxin. J. Bacteriol. 1965, 90 (4), 903–10. (33) Kittelberger, R.; Hilbink, F. Sensitive silver-staining detection of bacterial lipopolysaccharides in polyacrylamide gels. J. Biochem. Biophys. Methods 1993, 26 (1), 81–6. (34) Karkhanis, Y. D.; Zeltner, J. Y.; Jackson, J. J.; Carlo, D. J. A new and improved microassay to determine 2-keto-3-deoxyoctonate in lipopolysaccharide of Gram-negative bacteria. Anal. Biochem. 1978, 85 (2), 595–601. (35) Lamontagne, J.; Butler, H.; Chaves-Olarte, E.; Hunter, J.; Schirm, M.; Paquet, C.; Tian, M.; Kearney, P.; Hamaidi, L.; Chelsky, D.; Moriyon, I.; Moreno, E.; Paramithiotis, E. Extensive cell envelope modulation is associated with virulence in Brucella abortus. J. Proteome Res. 2007, 6 (4), 1519–29. (36) Lamontagne, J.; Beland, M.; Forest, A.; Cote-Martin, A.; Nassif, N.; Tomaki, F.; Moriyon, I.; Moreno, E.; Paramithiotis, E. Proteomicsbased confirmation of protein expression and correction of annotation errors in the Brucella abortus genome. BMC Genomics 2010, 11, 300. (37) Baldi, P.; Hatfield, G. W. DNA Microarrays and Gene Expression: From Experiments to Data Analysis and Modeling; Cambridge University Press: Oxford, 2002. (38) Sundaresh, S.; Doolan, D. L.; Hirst, S.; Mu, Y.; Unal, B.; Davies, D. H.; Felgner, P. L.; Baldi, P. Identification of humoral immune responses in protein microarrays using DNA microarray data analysis techniques. Bioinformatics 2006, 22 (14), 1760–6. (39) Sundaresh, S.; Randall, A.; Unal, B.; Petersen, J. M.; Belisle, J. T.; Hartley, M. G.; Duffield, M.; Titball, R. W.; Davies, D. H.; Felgner, P. L.; Baldi, P. From protein microarrays to diagnostic antigen discovery: a study of the pathogen Francisella tularensis. Bioinformatics 2007, 23 (13), i508–18. (40) Durbin, B. P.; Hardin, J. S.; Hawkins, D. M.; Rocke, D. M. A variance-stabilizing transformation for gene-expression microarray data. Bioinformatics 2002, 18 (Suppl 1), S105–10.

ARTICLE

(41) Huber, W.; von Heydebreck, A.; Sultmann, H.; Poustka, A.; Vingron, M. Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 2002, 18 (Suppl 1), S96–104. (42) Kreil, D. P.; Karp, N. A.; Lilley, K. S. DNA microarray normalization methods can remove bias from differential protein expression analysis of 2D difference gel electrophoresis results. Bioinformatics 2004, 20 (13), 2026–34. (43) Baldi, P.; Brunak, S. R. Bioinformatics: the machine learning approach; MIT Press: Cambridge, MA, 2001. (44) Long, A. D.; Mangalam, H. J.; Chan, B. Y.; Tolleri, L.; Hatfield, G. W.; Baldi, P. Improved statistical inference from DNA microarray data using analysis of variance and a Bayesian statistical framework. Analysis of global gene expression in Escherichia coli K12. J. Biol. Chem. 2001, 276 (23), 19937–44. (45) Benjamini, Y.; Hochberg, Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. Roy. Statist. Soc., Ser. B 1995, 57 (1), 289–300. (46) Moller, S.; Croning, M. D.; Apweiler, R. Evaluation of methods for the prediction of membrane spanning regions. Bioinformatics 2001, 17 (7), 646–53. (47) Bendtsen, J. D.; Nielsen, H.; von Heijne, G.; Brunak, S. Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 2004, 340 (4), 783–95. (48) Gardy, J. L.; Laird, M. R.; Chen, F.; Rey, S.; Walsh, C. J.; Ester, M.; Brinkman, F. S. PSORTb v.2.0: expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis. Bioinformatics 2005, 21 (5), 617–23. (49) Girosi, F. An Equivalence Between Sparse Approximation and Support Vector Machines. Neural Comput. 1998, 10 (6), 1455–80. (50) Ciftci, C.; Ozturk, F.; Oztekin, A.; Karaoglan, H.; Saba, R.; Gultekin, M.; Mamikoglu, L. Comparison of the serological tests used for the laboratory diagnosis of brucellosis. Mikrobiyol. Bul. 2005, 39 (3), 291–9. (51) Sisirak, M.; Hukic, M. Evaluation and importance of selected microbiological methods in the diagnosis of human brucellosis. Bosn. J. Basic Med. Sci. 2009, 9 (3), 198–203. (52) Dasgupta, A.; Sureka, K.; Mitra, D.; Saha, B.; Sanyal, S.; Das, A. K.; Chakrabarti, P.; Jackson, M.; Gicquel, B.; Kundu, M.; Basu, J. An oligopeptide transporter of Mycobacterium tuberculosis regulates cytokine release and apoptosis of infected macrophages. PLoS One 2010, 5 (8), e12225. (53) Skendros, P.; Pappas, G.; Boura, P. Cell-mediated immunity in human brucellosis. Microbes Infect. 2011, 13 (2), 134–42. (54) Paschos, A.; den Hartigh, A.; Smith, M. A.; Atluri, V. L.; Sivanesan, D.; Tsolis, R. M.; Baron, C. An in vivo high-throughput screening approach targeting the type IV secretion system component VirB8 identified inhibitors of Brucella abortus 2308 proliferation. Infect. Immun. 2011, 79 (3), 1033–43. (55) Wick, M. J.; Pfeifer, J. D.; Findlay, K. A.; Harding, C. V.; Normark, S. J. Compartmentalization of defined epitopes expressed in Escherichia coli has only a minor influence on efficiency of phagocytic processing for presentation by class I and class II major histocompatibility complex molecules to T cells. Infect. Immun. 1993, 61 (11), 4848–56. (56) Salih-Alj Debbarh, H.; Cloeckaert, A.; Bezard, G.; Dubray, G.; Zygmunt, M. S. Enzyme-linked immunosorbent assay with partially purified cytosoluble 28-kilodalton protein for serological differentiation between Brucella melitensis-infected and B. melitensis Rev.1-vaccinated sheep. Clin. Diagn. Lab. Immunol. 1996, 3 (3), 305–8. (57) Tibor, A.; Saman, E.; de Wergifosse, P.; Cloeckaert, A.; Limet, J. N.; Letesson, J. J. Molecular characterization, occurrence, and immunogenicity in infected sheep and cattle of two minor outer membrane proteins of Brucella abortus. Infect. Immun. 1996, 64 (1), 100–7. (58) Domingo, G. J.; Caivano, A.; Sartorius, R.; Barba, P.; Backstrom, M.; Piatier-Tonneau, D.; Guardiola, J.; De Berardinis, P.; Perham, R. N. Induction of specific T-helper and cytolytic responses to epitopes displayed on a virus-like protein scaffold derived from the pyruvate 4823

dx.doi.org/10.1021/pr200619r |J. Proteome Res. 2011, 10, 4813–4824

Journal of Proteome Research

ARTICLE

dehydrogenase multienzyme complex. Vaccine 2003, 21 (1314), 1502–9. (59) Magnan, C. N.; Zeller, M.; Kayala, M. A.; Vigil, A.; Randall, A.; Felgner, P. L.; Baldi, P. High-throughput prediction of protein antigenicity using protein microarray data. Bioinformatics 2010, 26 (23), 2936–43. (60) Corbin, R. W.; Paliy, O.; Yang, F.; Shabanowitz, J.; Platt, M.; Lyons, C. E., Jr.; Root, K.; McAuliffe, J.; Jordan, M. I.; Kustu, S.; Soupene, E.; Hunt, D. F. Toward a protein profile of Escherichia coli: comparison to its transcription profile. Proc. Natl. Acad. Sci. U.S.A. 2003, 100 (16), 9232–7. (61) Ishihama, Y.; Oda, Y.; Tabata, T.; Sato, T.; Nagasu, T.; Rappsilber, J.; Mann, M. Exponentially modified protein abundance index (emPAI) for estimation of absolute protein amount in proteomics by the number of sequenced peptides per protein. Mol. Cell. Proteomics 2005, 4 (9), 1265–72. (62) Ishihama, Y.; Schmidt, T.; Rappsilber, J.; Mann, M.; Hartl, F. U.; Kerner, M. J.; Frishman, D. Protein abundance profiling of the Escherichia coli cytosol. BMC Genomics 2008, 9, 102. (63) Kohler, S.; Foulongne, V.; Ouahrani-Bettache, S.; Bourg, G.; Teyssier, J.; Ramuz, M.; Liautard, J. P. The analysis of the intramacrophagic virulome of Brucella suis deciphers the environment encountered by the pathogen inside the macrophage host cell. Proc. Natl. Acad. Sci. U.S.A. 2002, 99 (24), 15711–6. (64) Bashir, A.; Bansal, V.; Bafna, V. Designing deep sequencing experiments: detecting structural variation and estimating transcript abundance. BMC Genomics 2010, 11, 385.

4824

dx.doi.org/10.1021/pr200619r |J. Proteome Res. 2011, 10, 4813–4824